summaryrefslogtreecommitdiff
path: root/Build/source/libs/harfbuzz/harfbuzz-src/src/gen-use-table.py
blob: 029e66e53a4194c8bebed1cd99d2750530531180 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
#!/usr/bin/env python
# flake8: noqa

from __future__ import print_function, division, absolute_import

import io
import sys

if len (sys.argv) != 5:
	print ("usage: ./gen-use-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt UnicodeData.txt Blocks.txt", file=sys.stderr)
	sys.exit (1)

BLACKLISTED_BLOCKS = ["Thai", "Lao"]

files = [io.open (x, encoding='utf-8') for x in sys.argv[1:]]

headers = [[f.readline () for i in range (2)] for j,f in enumerate(files) if j != 2]
headers.append (["UnicodeData.txt does not have a header."])

data = [{} for f in files]
values = [{} for f in files]
for i, f in enumerate (files):
	for line in f:

		j = line.find ('#')
		if j >= 0:
			line = line[:j]

		fields = [x.strip () for x in line.split (';')]
		if len (fields) == 1:
			continue

		uu = fields[0].split ('..')
		start = int (uu[0], 16)
		if len (uu) == 1:
			end = start
		else:
			end = int (uu[1], 16)

		t = fields[1 if i != 2 else 2]

		for u in range (start, end + 1):
			data[i][u] = t
		values[i][t] = values[i].get (t, 0) + end - start + 1

defaults = ('Other', 'Not_Applicable', 'Cn', 'No_Block')

# TODO Characters that are not in Unicode Indic files, but used in USE
data[0][0x034F] = defaults[0]
data[0][0x2060] = defaults[0]
# TODO https://github.com/roozbehp/unicode-data/issues/9
data[0][0x11C44] = 'Consonant_Placeholder'
data[0][0x11C45] = 'Consonant_Placeholder'
# TODO https://github.com/harfbuzz/harfbuzz/pull/1399
data[0][0x111C8] = 'Consonant_Placeholder'
for u in range (0xFE00, 0xFE0F + 1):
	data[0][u] = defaults[0]

# Merge data into one dict:
for i,v in enumerate (defaults):
	values[i][v] = values[i].get (v, 0) + 1
combined = {}
for i,d in enumerate (data):
	for u,v in d.items ():
		if i >= 2 and not u in combined:
			continue
		if not u in combined:
			combined[u] = list (defaults)
		combined[u][i] = v
combined = {k:v for k,v in combined.items() if v[3] not in BLACKLISTED_BLOCKS}
data = combined
del combined
num = len (data)


property_names = [
	# General_Category
	'Cc', 'Cf', 'Cn', 'Co', 'Cs', 'Ll', 'Lm', 'Lo', 'Lt', 'Lu', 'Mc',
	'Me', 'Mn', 'Nd', 'Nl', 'No', 'Pc', 'Pd', 'Pe', 'Pf', 'Pi', 'Po',
	'Ps', 'Sc', 'Sk', 'Sm', 'So', 'Zl', 'Zp', 'Zs',
	# Indic_Syllabic_Category
	'Other',
	'Bindu',
	'Visarga',
	'Avagraha',
	'Nukta',
	'Virama',
	'Pure_Killer',
	'Invisible_Stacker',
	'Vowel_Independent',
	'Vowel_Dependent',
	'Vowel',
	'Consonant_Placeholder',
	'Consonant',
	'Consonant_Dead',
	'Consonant_With_Stacker',
	'Consonant_Prefixed',
	'Consonant_Preceding_Repha',
	'Consonant_Succeeding_Repha',
	'Consonant_Subjoined',
	'Consonant_Medial',
	'Consonant_Final',
	'Consonant_Head_Letter',
	'Consonant_Initial_Postfixed',
	'Modifying_Letter',
	'Tone_Letter',
	'Tone_Mark',
	'Gemination_Mark',
	'Cantillation_Mark',
	'Register_Shifter',
	'Syllable_Modifier',
	'Consonant_Killer',
	'Non_Joiner',
	'Joiner',
	'Number_Joiner',
	'Number',
	'Brahmi_Joining_Number',
	# Indic_Positional_Category
	'Not_Applicable',
	'Right',
	'Left',
	'Visual_Order_Left',
	'Left_And_Right',
	'Top',
	'Bottom',
	'Top_And_Bottom',
	'Top_And_Right',
	'Top_And_Left',
	'Top_And_Left_And_Right',
	'Bottom_And_Left',
	'Bottom_And_Right',
	'Top_And_Bottom_And_Right',
	'Overstruck',
]

try:
	basestring
except NameError:
	basestring = str

class PropertyValue(object):
	def __init__(self, name_):
		self.name = name_
	def __str__(self):
		return self.name
	def __eq__(self, other):
		return self.name == (other if isinstance(other, basestring) else other.name)
	def __ne__(self, other):
		return not (self == other)
	def __hash__(self):
		return hash(str(self))

property_values = {}

for name in property_names:
	value = PropertyValue(name)
	assert value not in property_values
	assert value not in globals()
	property_values[name] = value
globals().update(property_values)


def is_BASE(U, UISC, UGC):
	return (UISC in [Number, Consonant, Consonant_Head_Letter,
			#SPEC-DRAFT Consonant_Placeholder,
			Tone_Letter,
			Vowel_Independent #SPEC-DRAFT
			] or
		(UGC == Lo and UISC in [Avagraha, Bindu, Consonant_Final, Consonant_Medial,
					Consonant_Subjoined, Vowel, Vowel_Dependent]))
def is_BASE_IND(U, UISC, UGC):
	#SPEC-DRAFT return (UISC in [Consonant_Dead, Modifying_Letter] or UGC == Po)
	return (UISC in [Consonant_Dead, Modifying_Letter] or
		(UGC == Po and not U in [0x104B, 0x104E, 0x2022, 0x111C8, 0x11A3F, 0x11A45, 0x11C44, 0x11C45]) or
		False # SPEC-DRAFT-OUTDATED! U == 0x002D
		)
def is_BASE_NUM(U, UISC, UGC):
	return UISC == Brahmi_Joining_Number
def is_BASE_OTHER(U, UISC, UGC):
	if UISC == Consonant_Placeholder: return True #SPEC-DRAFT
	#SPEC-DRAFT return U in [0x00A0, 0x00D7, 0x2015, 0x2022, 0x25CC, 0x25FB, 0x25FC, 0x25FD, 0x25FE]
	return U in [0x2015, 0x2022, 0x25FB, 0x25FC, 0x25FD, 0x25FE]
def is_CGJ(U, UISC, UGC):
	return U == 0x034F
def is_CONS_FINAL(U, UISC, UGC):
	# Consonant_Initial_Postfixed is new in Unicode 11; not in the spec.
	return ((UISC == Consonant_Final and UGC != Lo) or
		UISC == Consonant_Initial_Postfixed or
		UISC == Consonant_Succeeding_Repha)
def is_CONS_FINAL_MOD(U, UISC, UGC):
	#SPEC-DRAFT return  UISC in [Consonant_Final_Modifier, Syllable_Modifier]
	return  UISC == Syllable_Modifier
def is_CONS_MED(U, UISC, UGC):
	return UISC == Consonant_Medial and UGC != Lo
def is_CONS_MOD(U, UISC, UGC):
	return UISC in [Nukta, Gemination_Mark, Consonant_Killer]
def is_CONS_SUB(U, UISC, UGC):
	#SPEC-DRAFT return UISC == Consonant_Subjoined
	return UISC == Consonant_Subjoined and UGC != Lo
def is_CONS_WITH_STACKER(U, UISC, UGC):
	return UISC == Consonant_With_Stacker
def is_HALANT(U, UISC, UGC):
	return UISC in [Virama, Invisible_Stacker] and not is_HALANT_OR_VOWEL_MODIFIER(U, UISC, UGC)
def is_HALANT_OR_VOWEL_MODIFIER(U, UISC, UGC):
	# https://github.com/harfbuzz/harfbuzz/issues/1102
	# https://github.com/harfbuzz/harfbuzz/issues/1379
	return U in [0x11046, 0x1134D]
def is_HALANT_NUM(U, UISC, UGC):
	return UISC == Number_Joiner
def is_ZWNJ(U, UISC, UGC):
	return UISC == Non_Joiner
def is_ZWJ(U, UISC, UGC):
	return UISC == Joiner
def is_Word_Joiner(U, UISC, UGC):
	return U == 0x2060
def is_OTHER(U, UISC, UGC):
	#SPEC-OUTDATED return UGC == Zs # or any other SCRIPT_COMMON characters
	return (UISC == Other
		and not is_SYM_MOD(U, UISC, UGC)
		and not is_CGJ(U, UISC, UGC)
		and not is_Word_Joiner(U, UISC, UGC)
		and not is_VARIATION_SELECTOR(U, UISC, UGC)
	)
def is_Reserved(U, UISC, UGC):
	return UGC == 'Cn'
def is_REPHA(U, UISC, UGC):
	return UISC in [Consonant_Preceding_Repha, Consonant_Prefixed]
def is_SYM(U, UISC, UGC):
	if U == 0x25CC: return False #SPEC-DRAFT
	#SPEC-DRAFT return UGC in [So, Sc] or UISC == Symbol_Letter
	return UGC in [So, Sc]
def is_SYM_MOD(U, UISC, UGC):
	return U in [0x1B6B, 0x1B6C, 0x1B6D, 0x1B6E, 0x1B6F, 0x1B70, 0x1B71, 0x1B72, 0x1B73]
def is_VARIATION_SELECTOR(U, UISC, UGC):
	return 0xFE00 <= U <= 0xFE0F
def is_VOWEL(U, UISC, UGC):
	# https://github.com/roozbehp/unicode-data/issues/6
	return (UISC == Pure_Killer or
		(UGC != Lo and UISC in [Vowel, Vowel_Dependent] and U not in [0xAA29]))
def is_VOWEL_MOD(U, UISC, UGC):
	# https://github.com/roozbehp/unicode-data/issues/6
	return (UISC in [Tone_Mark, Cantillation_Mark, Register_Shifter, Visarga] or
		(UGC != Lo and (UISC == Bindu or U in [0xAA29])))

use_mapping = {
	'B':	is_BASE,
	'IND':	is_BASE_IND,
	'N':	is_BASE_NUM,
	'GB':	is_BASE_OTHER,
	'CGJ':	is_CGJ,
	'F':	is_CONS_FINAL,
	'FM':	is_CONS_FINAL_MOD,
	'M':	is_CONS_MED,
	'CM':	is_CONS_MOD,
	'SUB':	is_CONS_SUB,
	'CS':	is_CONS_WITH_STACKER,
	'H':	is_HALANT,
	'HVM':	is_HALANT_OR_VOWEL_MODIFIER,
	'HN':	is_HALANT_NUM,
	'ZWNJ':	is_ZWNJ,
	'ZWJ':	is_ZWJ,
	'WJ':	is_Word_Joiner,
	'O':	is_OTHER,
	'Rsv':	is_Reserved,
	'R':	is_REPHA,
	'S':	is_SYM,
	'SM':	is_SYM_MOD,
	'VS':	is_VARIATION_SELECTOR,
	'V':	is_VOWEL,
	'VM':	is_VOWEL_MOD,
}

use_positions = {
	'F': {
		'Abv': [Top],
		'Blw': [Bottom],
		'Pst': [Right],
	},
	'M': {
		'Abv': [Top],
		'Blw': [Bottom, Bottom_And_Left],
		'Pst': [Right],
		'Pre': [Left],
	},
	'CM': {
		'Abv': [Top],
		'Blw': [Bottom],
	},
	'V': {
		'Abv': [Top, Top_And_Bottom, Top_And_Bottom_And_Right, Top_And_Right],
		'Blw': [Bottom, Overstruck, Bottom_And_Right],
		'Pst': [Right, Top_And_Left, Top_And_Left_And_Right, Left_And_Right],
		'Pre': [Left],
	},
	'VM': {
		'Abv': [Top],
		'Blw': [Bottom, Overstruck],
		'Pst': [Right],
		'Pre': [Left],
	},
	'SM': {
		'Abv': [Top],
		'Blw': [Bottom],
	},
	'H': None,
	'HVM': None,
	'B': None,
	'FM': None,
	'SUB': None,
}

def map_to_use(data):
	out = {}
	items = use_mapping.items()
	for U,(UISC,UIPC,UGC,UBlock) in data.items():

		# Resolve Indic_Syllabic_Category

		# TODO: These don't have UISC assigned in Unicode 12.0, but have UIPC
		if 0x1CE2 <= U <= 0x1CE8: UISC = Cantillation_Mark

		# Tibetan:
		# TODO: These don't have UISC assigned in Unicode 12.0, but have UIPC
		if 0x0F18 <= U <= 0x0F19 or 0x0F3E <= U <= 0x0F3F: UISC = Vowel_Dependent
		if 0x0F86 <= U <= 0x0F87: UISC = Tone_Mark
		# Overrides to allow NFC order matching syllable
		# https://github.com/harfbuzz/harfbuzz/issues/1012
		if UBlock == 'Tibetan' and is_VOWEL (U, UISC, UGC):
			if UIPC == Top:
				UIPC = Bottom

		# TODO: https://github.com/harfbuzz/harfbuzz/pull/982
		# also  https://github.com/harfbuzz/harfbuzz/issues/1012
		if UBlock == 'Chakma' and is_VOWEL (U, UISC, UGC):
			if UIPC == Top:
				UIPC = Bottom
			elif UIPC == Bottom:
				UIPC = Top

		# TODO: https://github.com/harfbuzz/harfbuzz/pull/627
		if 0x1BF2 <= U <= 0x1BF3: UISC = Nukta; UIPC = Bottom

		# TODO: U+1CED should only be allowed after some of
		# the nasalization marks, maybe only for U+1CE9..U+1CF1.
		if U == 0x1CED: UISC = Tone_Mark

		# TODO: https://github.com/harfbuzz/harfbuzz/issues/525
		if U == 0x1A7F: UISC = Consonant_Final

		# TODO: https://github.com/harfbuzz/harfbuzz/issues/1105
		if U == 0x11134: UISC = Gemination_Mark

		# TODO: https://github.com/harfbuzz/harfbuzz/pull/1399
		if U == 0x111C9: UISC = Consonant_Final

		values = [k for k,v in items if v(U,UISC,UGC)]
		assert len(values) == 1, "%s %s %s %s" % (hex(U), UISC, UGC, values)
		USE = values[0]

		# Resolve Indic_Positional_Category

		# TODO: These should die, but have UIPC in Unicode 12.0
		if U in [0x953, 0x954]: UIPC = Not_Applicable

		# TODO: In USE's override list but not in Unicode 12.0
		if U == 0x103C: UIPC = Left

		# TODO: These are not in USE's override list that we have, nor are they in Unicode 12.0
		if 0xA926 <= U <= 0xA92A: UIPC = Top
		# TODO: https://github.com/harfbuzz/harfbuzz/pull/1037
		#  and https://github.com/harfbuzz/harfbuzz/issues/1631
		if U in [0x11302, 0x11303, 0x114C1]: UIPC = Top
		if U == 0x1171E: UIPC = Left
		if 0x1CF8 <= U <= 0x1CF9: UIPC = Top

		assert (UIPC in [Not_Applicable, Visual_Order_Left] or
			USE in use_positions), "%s %s %s %s %s" % (hex(U), UIPC, USE, UISC, UGC)

		pos_mapping = use_positions.get(USE, None)
		if pos_mapping:
			values = [k for k,v in pos_mapping.items() if v and UIPC in v]
			assert len(values) == 1, "%s %s %s %s %s %s" % (hex(U), UIPC, USE, UISC, UGC, values)
			USE = USE + values[0]

		out[U] = (USE, UBlock)
	return out

defaults = ('O', 'No_Block')
data = map_to_use(data)

print ("/* == Start of generated table == */")
print ("/*")
print (" * The following table is generated by running:")
print (" *")
print (" *   ./gen-use-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt UnicodeData.txt Blocks.txt")
print (" *")
print (" * on files with these headers:")
print (" *")
for h in headers:
	for l in h:
		print (" * %s" % (l.strip()))
print (" */")
print ()
print ('#include "hb-ot-shape-complex-use.hh"')
print ()

total = 0
used = 0
last_block = None
def print_block (block, start, end, data):
	global total, used, last_block
	if block and block != last_block:
		print ()
		print ()
		print ("  /* %s */" % block)
		if start % 16:
			print (' ' * (20 + (start % 16 * 6)), end='')
	num = 0
	assert start % 8 == 0
	assert (end+1) % 8 == 0
	for u in range (start, end+1):
		if u % 16 == 0:
			print ()
			print ("  /* %04X */" % u, end='')
		if u in data:
			num += 1
		d = data.get (u, defaults)
		print ("%6s," % d[0], end='')

	total += end - start + 1
	used += num
	if block:
		last_block = block

uu = sorted (data.keys ())

last = -100000
num = 0
offset = 0
starts = []
ends = []
print ('#pragma GCC diagnostic push')
print ('#pragma GCC diagnostic ignored "-Wunused-macros"')
for k,v in sorted(use_mapping.items()):
	if k in use_positions and use_positions[k]: continue
	print ("#define %s	USE_%s	/* %s */" % (k, k, v.__name__[3:]))
for k,v in sorted(use_positions.items()):
	if not v: continue
	for suf in v.keys():
		tag = k + suf
		print ("#define %s	USE_%s" % (tag, tag))
print ('#pragma GCC diagnostic pop')
print ("")
print ("static const USE_TABLE_ELEMENT_TYPE use_table[] = {")
for u in uu:
	if u <= last:
		continue
	block = data[u][1]

	start = u//8*8
	end = start+1
	while end in uu and block == data[end][1]:
		end += 1
	end = (end-1)//8*8 + 7

	if start != last + 1:
		if start - last <= 1+16*3:
			print_block (None, last+1, start-1, data)
			last = start-1
		else:
			if last >= 0:
				ends.append (last + 1)
				offset += ends[-1] - starts[-1]
			print ()
			print ()
			print ("#define use_offset_0x%04xu %d" % (start, offset))
			starts.append (start)

	print_block (block, start, end, data)
	last = end
ends.append (last + 1)
offset += ends[-1] - starts[-1]
print ()
print ()
occupancy = used * 100. / total
page_bits = 12
print ("}; /* Table items: %d; occupancy: %d%% */" % (offset, occupancy))
print ()
print ("USE_TABLE_ELEMENT_TYPE")
print ("hb_use_get_category (hb_codepoint_t u)")
print ("{")
print ("  switch (u >> %d)" % page_bits)
print ("  {")
pages = set([u>>page_bits for u in starts+ends])
for p in sorted(pages):
	print ("    case 0x%0Xu:" % p)
	for (start,end) in zip (starts, ends):
		if p not in [start>>page_bits, end>>page_bits]: continue
		offset = "use_offset_0x%04xu" % start
		print ("      if (hb_in_range<hb_codepoint_t> (u, 0x%04Xu, 0x%04Xu)) return use_table[u - 0x%04Xu + %s];" % (start, end-1, start, offset))
	print ("      break;")
	print ("")
print ("    default:")
print ("      break;")
print ("  }")
print ("  return USE_O;")
print ("}")
print ()
for k in sorted(use_mapping.keys()):
	if k in use_positions and use_positions[k]: continue
	print ("#undef %s" % k)
for k,v in sorted(use_positions.items()):
	if not v: continue
	for suf in v.keys():
		tag = k + suf
		print ("#undef %s" % tag)
print ()
print ("/* == End of generated table == */")

# Maintain at least 50% occupancy in the table */
if occupancy < 50:
	raise Exception ("Table too sparse, please investigate: ", occupancy)